Vuln Processing crashes 4 GB container after upgrading to v4.90.1

[rfairburn]

Fleet versions

• Discovered: v4.90.1
• Reproduced: N/A (not yet reproduced on a clean install)

Web browser and operating system: N/A (server-side issue, observed via container crash logs)

---

💥 Actual behavior

Within a couple of minutes of a vuln processing run starting after upgrading to v4.90.1, the containers were crashing with an out-of-memory error. This was recorded on a number of customer environments as outlined by the attached labels.

Increasing memory to 8 GB appears to have helped for customer-traino, but I don't have retry data for the others yet.

🛠️ Expected behavior

Vuln processing completes without crashing the containers under the standard 4 GB memory limit.

🧑‍💻 Steps to reproduce

These steps:

• Have been confirmed to consistently lead to reproduction in multiple Fleet instances.
• Describe the workflow that led to the error, but have not yet been reproduced in multiple Fleet instances.

1. Upgrade an existing installation to v4.90.1.
2. Allow a vuln processing run to start.
3. Observe container OOM crash within a couple of minutes.

🕯️ More info (optional)

• Affected customer environments are indicated by the attached customer-* labels.
• Raising the container memory limit to 8 GB appears to mitigate the issue (confirmed so far for one environment; retry data pending for the rest).

[rfairburn]

Update: customer-numa is also experiencing and hasn't updated to v4.90.1 yet! It might be something with the actual vuln data has changed. Confirmed v4.89.2-v4.90.1 is the detected range.